Ebbe's Keynote Critiques

[Medhue Simoni]

---

Vivienne Daguerre wrote:

 I am very excited by it and cannot wait to try it out! He did mention that there would be some form of in world build tool.

"Support for voxels is under consideration with the next generation platform, to present an in-world option for content creation, terrain modification, etc."

"Like I said, over time, we’re obviously going to make it easier to do layout without within the world, but we’re also exploring technologies like voxels to think of ways to make it easy for non-3D experts to be able to create environments and structures; so that’s an area we’re investing some time in right now, to understand what we can bring to the table there."

"So we can hit a much broader range of creators,  from professionals who can use the tools they’re comfortable with today to hobbyists who are willing to learn some new tools and who could benefit from using things like voxel systems to easily “paint”  and chip away to create terrain and tunnels and caves and stuff like that."

As I understand it, the Next Gen Platform is looking at voxel based terrain, which is a wonderful thing! It means you can dig or blast holes in it, create caves and tunnels, and paint texture or colour directly onto the terrain. Imagine the possibilities of that!

---

Well, the whole voxel thing is interesting, and I'll be happy if the new world has some inworld tools. It definitely opens up possibilities, although I'll likely not play with it much. To me, and this is probably because of what my specialty is, the biggest factor that worries me is the avatar. There is no more important factor than the skeleton, and avatar features. The reason I worry, is because LL has shown little to no interest in Avatars. From what LL shows us, they don't even have someone employed by LL who is an avatar expert, or even anyone who can make a mesh. I find that just completely crazy. They have experts right at their finger tips, in Gaia and the Machinimatrix team. Why is LL not working directly with them? Why isn't LL throwing money at them, cause I would be. Everything else in the new world is secondary to the avatar. If they get that wrong, or do it bad, it really doesn't matter what they do with the rest, IMHO.

And personally, I think it is somewhat wrong to put alot of focus on Oculus or VR stuff. Of course you want it all to work, but I think headsets like that are just alot of hype. How many times have we heard how 3D is the next big thing, to the point companies spend millions on 3D TV technology, and nobody really cares about it. It might be better in the future, but it's not mainstream. The same goes for VR headsets. It's just not there yet. I think before masses ever put something on their heads, we'll be projecting holographic worlds into our living rooms. People don't like to wear stuff, especially on their heads. Even myself, as an animator, I like camera based tracking much better than sensor tracking, plus it's alot less hardware.

Heck, just look at Google glasses. Who actually uses that? It was interesting, but beyond that, not all that practical, outside of some small use cases.

 

---

Vivienne Daguerre wrote:

I would like to steer the discussion away from political science and "Linden Lab is malicious and evil," back to a discussion of the next generation platform discussed by Ebbe.

---

I don't know if others think LL is malicious or evil, but I never said or implied that. I've always talked directly about their management, interests, and competence/enthusiasm for a project. For instance, I complained about fitted mesh and it's many faults. That has nothing at all to do with maliciousness or evil intent. As a content creator tho, it's a nightmare for us. It truly is, and a ton of needless work.

[Vivienne Daguerre]

I agree that the default avatar will be important, if there is one, and I am not sure there will be. High Fidelity is open alpha now, so we can talk about it a little. I don't think they intend to have a default avatar. You will upload something you made, made by someone else, or buy an avatar that you like in world from someone else. You will then have to buy clothing made for that specific avatar from the maker of it, or designers working with it. Favourites will rise to the top, and there will be great variety. 

I suspect that there will be custom avatars, like the mesh avatars available in SL now, and a default avatar of some sort. I do think that LL will go to someone else to make them, as they did with the new default mesh avatars (Nylong Pinkney I think). 

I hope that they will enable custom skeletons for animals and npcs. I loved your video of your elephant, and hope this will be possible in SL. I made animals in SL in pieces and painfully scripted the parts to move to form animations. It was very awkward.

You never said that SL was malicious or evil, but in the thread there were some of the usual suspects who espoused that point of view, which is not helpful in a discussion.

[LlazarusLlong]

---

Vivienne Daguerre wrote:

 I don't think they intend to have a default avatar. You will upload something you made, made by someone else, or buy an avatar that you like in world from someone else. You will then have to buy clothing made for that specific avatar from the maker of it, or designers working with it. Favourites will rise to the top, and there will be great variety.

---

A recipe for disaster!

Supply led markets didn't work for Communism, why should they work in virtual capitalism.

[Medhue Simoni]

---

Vivienne Daguerre wrote:

I agree that the default avatar will be important, if there is one, and I am not sure there will be. High Fidelity is open alpha now, so we can talk about it a little. I don't think they intend to have a default avatar. You will upload something you made, made by someone else, or buy an avatar that you like in world from someone else. You will then have to buy clothing made for that specific avatar from the maker of it, or designers working with it. Favourites will rise to the top, and there will be great variety. 

I suspect that there will be custom avatars, like the mesh avatars available in SL now, and a default avatar of some sort. I do think that LL will go to someone else to make them, as they did with the new default mesh avatars (Nylong Pinkney I think).

---

I've been in HF, and I've looked at their avatar. As far as I understand it, you can make any avatar, but to have them work with everything, then they have to be made a specific way, with specific bone names. This is especially important if your avatar is going to use the Face controls with a 3D camera. I haven't actually uploaded anything there yet, as they seem a bit too far off for me to dig in yet. It's kind of tough for me to spend all that much time just testing it, with no return coming anytime soon. And.... actually just yesterday, I was reading how HF used Maya, and now the Blender users can't get their rigs to work, and it looks like almost the exact issue that Cloud Parties rig had. Seriously, does anyone want to bet me that LL's new world will not have the exact same issue?

I like the idea of a default avatar. It allows for many things. Heck, for something like SL, you have to have a default rig, because how else could a good animation market form, where the animations work for every humanoid avatar? In Unity, they have a humanoid default rig, and as long as you have a similar bones set up as their default, you can use both your own rig, and their humanoid rig, so your avatar can basically use 2 rigs independently, with the humanoid default rig handling all the main body movements, and your rig handling anything outside of the humanoid rig. It can be pretty nifty, especailly for facial rigs, tails and ears and stuff.

The other part about why a default is good, is the geometry. You'll have 1 consistent geometry for the avatar. Not 1 with millions of triangles, next to an efficiently made avatar. With everyone using the same avatar, and the same morphs, for clothing and other accessories, this would standardize things. What LL should do tho, is allow us to import our own morphs or blendshapes for their default avatar. This should be easy, because as long as the mesh avatar you upload has the same geometry as the default avatar, it should be easy for LL to create a morph slider from that mesh. So, someone could take the default avatar mesh, bring it into Blender or Zbrush, and sculpt a new shape out of it. They just can't change the geometry of the mesh. If LL did this, selling new morphs for avatars would be a huge new industry. It would be like Dazstudio's Genesis character, inwhich I easily have over 500 different morphs for it, or more.

[irihapeti]

---

Medhue Simoni wrote:

Everything else in the new world is secondary to the avatar. If they get that wrong, or do it bad, it really doesn't matter what they do with the rest, IMHO.

 

---

agree

[Coby Foden]

---

irihapeti wrote:

---

Medhue Simoni wrote:

Everything else in the new world is secondary to the avatar. If they get that wrong, or do it bad, it really doesn't matter what they do with the rest, IMHO.

---

agree

---

I agree too.

Just looking at how much people spend time, effort and money to beautify their avatars it's pretty obvious that most people agree - the avatar is the most important thing.

[Vivienne Daguerre]

Well said Medhue. After reading your post, I must agree with you on all points.

[LlazarusLlong]

It sounds like you lot would be happy with sophisticated avatars jumping about in front of a green screen on your monitor, with you individually deciding what image you wanted as a background.

Not interested in a completely shared experience then?

[Medhue Simoni]

---

LlazarusLlong wrote:

It sounds like you lot would be happy with sophisticated avatars jumping about in front of a green screen on your monitor, with you individually deciding what image you wanted as a background.

Not interested in a completely shared experience then?

---

I never said that, nor implied it. As a creator tho, I understand that terrain, and backgrounds stuff are all just basic modeling and texturing. This is not to say that it is easy to do really well, but that it's not really new. As much as voxels alter the process a little, it's still kind of the same. My point is that it's not that complex.

The avatar, on the other hand, has almost unlimited levels of complexity, and every single platform handles it differently.

[Medhue Simoni]

---

Vivienne Daguerre wrote:

Well said Medhue. After reading your post, I must agree with you on all points.

---

Thanks Vivienne!

LL should really consider allowing us to morph the default avatar. This is why I'm so annoyed with the closed nature of how LL is creating this new world. You have to actually be an avatar creator to understand the possibilities, and as far as I know, LL doesn't have a knowledgable paid advisor on this stuff. And..... to anyone thinking I just want a job...... LL would have to pay me quite a bit to pull me away from my own projects, and that is pretty implausable. I'm very willing to just comment freely on what I do know and see. If the process was wide open tho, then LL would have many different creators chiming in with good ideas. They should take advantage of that with open meetings and stuff. Many of us artists have seen many different ways of doing things. The first time I saw morphs as an actual traded element was in Daz, and the whole concept, from an arstist's and consumer's stand point, is fricken awesome.

 

When you are relying mostly on coders to think things up, you are going to get vastly different results than anything an artist would think up. I'm not saying coders aren't artists, cause they are, but a very different kind of art. I've worked with many, many coders now. Today, when a client wants me to do something for them, I now always demand that the coder and I work together to figure out how to do things. When I just do it how I think it should be done, or they just dictate to me how to do it, it never works out, or could have been done much better. We save a lot of money and time with just me and the coder talking it out, maybe doing some quick test to finalize how it should work.

Right now, I'm the lead on my own game. We have a 3 person team, of a coder, an artist, and me. Our main avatar is as complex as what LL needs to do with their new avatar, if not much more complicated. Our main character is an animal, and it needs to be able to morph into almost any animal. Working it all out is incredibly difficult and can be quite confusing. If the coder just coded the system herself, it would be vastly different than what it is becoming, even tho she does have some 3D creation knowledge. And actually, she said it can't be done. Of course, I know better, and now we are banging it out to make it happen.

 

 P.S. They should also let us animate those morphs/blendshapes in the new world. I'll admit, this is for completely selfish reasons initially, as I'm a big Werewolf fan. It would fricken rock to really morph people into werewolves in the new world. Plus, with animated blendshapes, you could simulate cloth physics, instead of actually processing real cloth physics. (oh, and I did mention this to Ebbe, in response to his response to me, on Drax's podcast comment page. Maybe he didn't see it tho)

[irihapeti]

---

Medhue Simoni wrote:

 

 Plus, with animated blendshapes, you could simulate cloth physics, instead of actually processing real cloth physics.

---

you really onto something with this

if you and your team can make this work then will be a pretty major step forward

is very well worth pursuing. It will solve a whole heap of problems with current tech and will also be lots of interest for a working algo/solution from  the big commercial players in it

so I encourage you to keep working on this path

 

[Medhue Simoni]

---

irihapeti wrote:

---

Medhue Simoni wrote:

 

 Plus, with animated blendshapes, you could simulate cloth physics, instead of actually processing real cloth physics.

---

you really onto something with this

if you and your team can make this work then will be a pretty major step forward

is very well worth pursuing. It will solve a whole heap of problems with current tech and will also be lots of interest for a working algo/solution from  the big commercial players in it

so I encourage you to keep working on this path

 

---

This is not really something that is new. Many games use it. In Watch Dogs, I'm pretty sure there is no real cloth physics in the game, but when you jump on a motorcycle, you will see his coat flaps in the wind. If you look closely, that flapping is always the same. I've also seen it in a few batman games, for his cape. If you don't create avatars tho, you likely won't even know how this is done.

In Blender, it's actually quite easy. You just set up your cloth physic in blender. Run it, and then goto the frame you like, and convert the shape into a blend shape.

[Derek Torvalar]

In case you may have missed this. Courtesy of Indigo Mertel.

http://blog.nalates.net/2015/04/18/next-generation-platform-sl2-the-avatars/

[Medhue Simoni]

---

Derek Torvalar wrote:

In case you may have missed this. Courtesy of Indigo Mertel.

http://blog.nalates.net/2015/04/18/next-generation-platform-sl2-the-avatars/

---

Thanks Derek! It's a good article. Here is my response. I posted it there, but I got an error, so I don't know if it will stick there.

 

It's not a bad description of the hurdles Nalates. Blend shapes, or morphs are a must have for NGP. There's also some pretty cool stuff creators could use blend shapes for, like faking real physics in hair or clothing, saving lots of processing of realtime cloth physics.

 

Allow me to throw in what I think is a decent concept to fix some of these solutions. I spent a fair bit of time playing with Daz3D's Genesis character, and making animations, as well as morphs for it. I am proposing that LL do something similar.  The Genesis character is a high poly model. I'm not saying LL's default needs to have so many polys, but it should be a decent amount for the platform LL is creating. Then, LL should allow us to import full body morphs of that default, or morph of just specific parts. Once the creator uploads the morph, and LL's system creates the slider, the creator now can sell that morph asset, just like you can for Genesis in DazStudio. A customer could buy the morph asset for their default avatar. Then, that customer could transfer the morph shape to clothing items.

 

To me, it just really makes sense to do things this way. Yeah, people could upload their own mesh avatar, but it will be beneficial to just morphs the default into the body you want as a creator, to have access to other features the default has, like animations too. Plus, with LL dictating the to polygons and typology of the default, overall, things will be more consistent, and easier for the average user to understand and know what works. This 1 feature, of morphing the default, could lead to some really crazy morphs created for it, which alter the avatar entirely. I easily have 500 - 1000 morphs for Genesis, which I've even converted over to Blender. It almost ridiculous how many morphs I have for him.

 

It would also be very easy to do for creators. You'd just bring the default mesh into Blender or Zbrush, and sculpt away on that mesh. For those not all that good at sculpting, you just use proportional editing to alter things, and move some verts around. Heck, in Blender, you could bring in a mesh with the head you did, then just use the snap to face feature to form to the head you like. Really there are endless possibilities.

[kiramanell]

“No more Linden Programming Language, so you can use existing experts.”

Wait, wut?! No more LSL?! Oh yeah, that's gonna work great for backwards compatibility.

[Dresden Ceriano]

---

kiramanell wrote:

“No more Linden Programming Language, so you can use existing experts.”

Wait, wut?! No more LSL?! Oh yeah, that's gonna work great for backwards compatibility.

---

Perhaps you missed the part where Ebbe said LL wasn't intending to concern themselves with backwards compatibility on the new platform.

...Dres

[Innula Zenovka]

---

kiramanell wrote:

“No more Linden Programming Language, so you can use existing experts.”

Wait, wut?! No more LSL?! Oh yeah, that's gonna work great for backwards compatibility.

---

LL have said from the outset that LSL won't be used in The New Thing.   That's a very sensible decision, to my mind.   C# is an obvious choice -- Unity and Unreal Engine use it already --   and I'm finding it quite easy to teach myself (my background is in business and law, not computing).   If anying, I think people will find it easier than LSL because it's a lot more versatile.   

Probably some enterprising person will write a set of tools to translate LSL into C# (I think people have discussed it already in various threads) but many scripts you probably wouldn't want to translate because the environment is so different.   Anything that moves, I'd think, is going to work far better if it's rescripted from scratch, for the new environment, rather than use something designed for SL.   

Sticking with LSL would have been a real missed opportunity.   I've never been very patient with people who complain that LSL doesn't do this or that thing that a "real" language does, because there's almost always work-rounds, but I'm very glad -- now I'm learning C# -- that I'll have a lot of very useful tools at my disposal that aren't available in LSL.

[Dresden Ceriano]

---

Medhue Simoni wrote:

Thanks Derek! It's a good article. Here is my response. I posted it there, but I got an error, so I don't know if it will stick there.

---

Go back and try again, Medhue.  I got an error the first two times I tried posting as well.  Only after I realized that I had to refresh the captcha device, before entering my answer into it, was I able to successfully post my reply.

...Dres

[Coby Foden]

---

kiramanell wrote:

“No more Linden Programming Language, so you can use existing experts.”

Wait, wut?! No more LSL?! Oh yeah, that's gonna work great for backwards compatibility.

---

The Next Generation Platform (as Linden Lab refers to it at the moment) will not be an upgrade of Second Life.

It is totally brand new world. Just think that it has nothing at all to do with SL and then you're fine.

[kiramanell]

---

Coby Foden wrote:

---

kiramanell wrote:

“No more Linden Programming Language, so you can use existing experts.”

Wait, wut?! No more LSL?! Oh yeah, that's gonna work great for backwards compatibility.

---

The

Next Generation Platform

(as Linden Lab refers to it at the moment) will

not

be an upgrade of Second Life.

It is totally brand new world. Just think that it has nothing at all to do with SL and then you're fine.

---

I had always entertained the (perhaps idle) hope that there would be sort of if not backwards compatibility, then at least some importability. But, as another poster said, obsolescence has become the new business model.

 

[Vivienne Daguerre]

I have been learning C# too. Having seen LSL, the structure was not too hard to pick up.  C# is more complex, but it did not seem completely foreign.

One friend said that he did not understand why or how they could permit full use of C#, because it would be possible to do too much with it, things that would affect the viewer interface and perhaps even do things to the user's computer. I am no expert, but I suspect they will likely limit the use to specific libraries. They may even provide a library of methods that we can use. 

(Methods would be equivalent to functions in LSL.)

The first lines of a C# program are to list the libraries used, and all the methods within those libraries then become available to the program. Here is an example:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

 

The Linq library has methods useful to handling collections, for example. In the next gen platform we may have a requirement to use a defined list of libraries, and be limited to those, or perhaps even just one, so it might look something like this:

using Linden.Lab.Methods;

As I say, I am no expert, and this is my guess at how it might be implemented. If I am right, we can learn basic C# and its syntax to be ready for the next gen platform, but once there, we may have to learn specific methods permitted to us. I don't think that will be a big deal.

[Madelaine McMasters]

The language isn't what determines how much you can do. As you've surmised, it's the library functions and system interfaces that set the bounds. And the scripting language won't be running on your computer, it'll be running in the virtual world's servers. 

Even if LL is contemplating a system in which scripts can run on the viewer, they'd be constrained within the bounds of the script interpreter. I imagine LL would provide the same sort of sandboxing that modern web browsers provide for Java/Javascript.

Your friend was right about not understanding. You may be more expert than he is!

;-).

[Vivienne Daguerre]

Thanks for that, helper from the future! :catwink:

[Deltango Vale]

I find it amazing how few people understand virtual worlds, but I'm baffled as to why so few people in Second Life understand virtual worlds. Btw, irihapeti, this is not directed against you personally.

When the first Model-T was produced, people gawked and laughed. There were lots of photos of cars bogged down in the mud or being outrun by a horse. Haha, who wants one of those silly contraptions? One spends 10 minutes driving and 10 hours fixing it.

So, easy target, you say. Okay, I'll give you a photo of a modern Mercedes Benz and put you in a time machine and send you back to show the folks what a car will become. They look at the photo and smirk. They don't believe a word of it. Why? Because they have a certain mentality shaped by their experiences of their era. They simply can't grasp the notion of independent electric motors powering flat assembly lines or fuel injection or airbags or microchips or robotics. They lack the mental framework and the million bits of knowledge and experience of the modern world. They look at the photo and see an artist's drawing for a science fiction magazine. Remember how people scoffed at Dick Tracy's watch? It's only now that people can conceive it working because only now do they have the understanding of how such a watch could work.

Same goes for virtual worlds. Most people don't have the mental framework or experience to see how virtual worlds will become the dominant political-economic structure of the 21st century. Okay, maybe I'm being a bit dramatic here, but I'm trying hard to make a point. Just because Second Life is the Model-T of virtual worlds doesn't mean one should dismiss it as a fad or something too complicated for the masses. Grandpa may have struggled with a mobile phone, buy your kids manage it in their sleep.

As for the SL economy, remember two things: 1) it is mostly residential and 2) RL is a source of money. We don't have the luxury in RL of paying the rent from a bank account on Jupiter. We do have that luxury in SL. Whip out a credit card and Bob's your uncle. Also, think about why people have cottages. They want to get away from their normal lives and relax at the cottage. SL is 'the cottage' for most residents.

All this adds up to a massive commercial opportunity that will make Facebook seem like 'Show & Tell'. Virtual worlds - and SL if it were managed right - will be FB x 1000. It's blindingly obvious to anyone who understands history and human nature.

[ZoeTick]

---

Deltango Vale wrote:

Second Life is the Model-T of virtual worlds

 

---

Hey Del, are you ABSOLUTELY sure that SL isn't actually the Edsel of virtual worlds?

It sounds more like it.